Predictive Intelligence: Issue with populating data in the Demand Insights DashboardIssue Sometimes, customers will create their own custom similarity solution and clustering solution and add these solutions to the demand insight configuration and experience that the data is not populating on the Demand insights Dashboard. They will add their own custom fields in the demand insight configuration for specific table. If the field types are journal list or string fields, there will not be any issue observed and data will be populated on the Demand insights dashboard as expected. If the field types used are reference fields, they will experience data is not populated on the Demand insights dashboardReleaseAllCauseMost Probable cause could be because of the reference field is not dot walked. If there is reference field for example " subject detail" , it should be dot walked like (subject detail.name or subject detail.id or subject detail.company)ResolutionCheck for the system log error "KBCuration :: Exception while checking similar articles for tasks java.lang.NullPointerException null undefined"Check for the fields that are used in the Demand insights configuration and ensure dot walking is done as expected for the reference fields used.Once dot walking is complete, you should find the data is populated on the Demand insights dashboard as expected.In case if configuration and fields are working as expected, the information message will be shown like below in the system log. KB curation started and KB curation finished, once these messages are displayed, the solutions will be trained with new versions and training will be completed. As soon as the training completes, the data will be populating on the dashboard as per the customer expectation.